1. Field of the Invention
This invention relates generally to the management of so-called contact spots through which, on a micro-scopic scale, electrical currents are conducted across interfaces of solids, whether between the two sides of switches or between sliding as well as stationary electrical brushes and their substrates, being mostly but not exclusively slip rings and commutator bars.
The electrical brushes at issue include fiber brushes disclosed in the above-noted U.S. Pat. Nos. 4,358,699 and 4,415,635, and in U.S. Pat. No. 6,245,440. Additionally, they include foil brushes as described in the publication xe2x80x9cProduction and Performance of Metal Foil Brushes,xe2x80x9d P. B. Haney, D. Kuhlmann-Wilsdorf and H. G. F. Wilsdorf, WEAR, 73 (1981), pp. 261-282, which is also incorporated by reference, and ordinary monolithic brushes made of graphite or graphite-metal mixtures. The invention is also applicable to electrical switches for the reduction of resistance and sticking forces, as well as to devices for efficient heat transfer.
The present invention includes the use of various technologies referenced and described in the above-noted U.S. Patents and Applications, as well as described in the references identified in the appended APPENDIX and cross-referenced throughout the specification by reference to the corresponding number, in brackets, of the respective references listed in the APPENDIX, the entire contents of which, including the related patents and applications listed above and the references listed in the APPENDIX, are incorporated herein by reference.
2. Discussion of the Background
Sliding electrical contacts, i.e., xe2x80x9cbrushesxe2x80x9d, conduct electrical current between solids, very preponderantly metals, in relative motion. Brushes are in widespread use in various types of electric motors and generators and are also widely used in less common but numerous special applications, e.g. telemetry devices and rotating antennae. Even while to date the traditional xe2x80x9cmonolithicxe2x80x9d (i.e., in the form of a solid piece) graphite-based (i.e., including compacted graphite or various metal-graphite mixtures) brushes are overwhelmingly frequent, they have a number of technological limitations. Specifically, monolithic graphite-based brushes cannot be reliably used, over extended periods of time, at current densities above about 30 Amp/cm2, nor at sliding speeds above about 25 m/sec. Further, as a coarse estimate, they waste about one watt per ampere conducted across the brush-substrate interface, i.e. the equivalent of one Volt, in terms of Joule and friction heat. Further, they emit significant intensities of electromagnetic waves (i.e., they are electrically very noisy so as to interfere with radio and similar signal reception), and finally they wear into a powdery debris that can be highly detrimental in electrical machinery, especially aboard submarines.
As a result of these shortcomings of traditional monolithic brushes, a number of otherwise very attractive technological developments are stymied for lack of electrical brushes which will conduct reliably over extended time periods, much higher current densities at low losses up to much higher speeds. Most importantly impacted are so-called xe2x80x9chomopolarxe2x80x9d motors and generators. They have potentially very high power densities and would be excellent for Navy ship drives, among others, but typically require current densities in excess of one hundred Amperes per cm2 to be conducted across interfaces of metal parts relatively moving at sustained speeds up to of 30 m/sec or even more while producing or requiring EMF""s of only 20V or so. The requirements of homopolar machinery in terms of current densities and speeds can thus not be fulfilled by monolithic brushes, and in any event a loss of 2 Volts per monolithic brush pair, i.e., in and out, is prohibitive for homopolar machines.
In previous inventions, particularly in the Patent Application xe2x80x9cContinuous Metal Fiber Brushes, [1]xe2x80x9d the capabilities of metal fiber brushes, including multitudes of essentially parallel hair-fine metal fibers, are outlined. They are intrinsically capable of easily conducting the desired current densities and to do so up to at least 70 m/sec with a total loss in the order of 0.1 Volt per brush. At the same time such brushes are electrically very quiet. These superior qualities derive from large numbers of separate electric xe2x80x9ccontact spotsxe2x80x9d, namely at the fiber ends at the brush xe2x80x9cworking surfacexe2x80x9d sliding along the brush-substrate interface, through which the current is physically conducted on a microscopic scale. That current is conducted across solid interfaces only through a restricted number of contact spots, whose total area amounts to only fractions of one percent of the macroscopic area of contact, is a well-known general physical phenomenon. To a large extent the poor qualities of monolithic brushes arise from their small number of contact spots, namely in the order of ten per brush. As a result, the current flow lines in monolithic brushes are not rather uniformly distributed, as they are in metal fiber brushes, but they are xe2x80x9cconstricted [2]xe2x80x9d at the few contact spots. This causes the corresponding xe2x80x9cconstriction resistancexe2x80x9d that represents in the order of one third the resistance of monolithic brushes.
The superiority of metal fiber brushes does not only derive from their thousands of evenly distributed contact spots, but also from the fact that at their contact spots bare metal meets bare metal, ideally separated only by a double monomolecular layer of adsorbed water vapor. Fortuitously, this most favorable type of lubrication, which prevents cold-welding and accommodates the relative motion between brush and substrate at a xe2x80x9cfilm resistivityxe2x80x9d of only "sgr"Fxe2x88x921xc3x9710xe2x88x9212 xcexa9m2 and average friction coefficient (xcexc) of about 0.3, establishes itself automatically at any modest ambient humidity, provided that undue contamination with oils, etc., is avoided. By contrast, monolithic brushes deposit a lubricating graphitic layer through which the current must flow at much higher electrical film resistivity. Further, the body resistance of graphitic brushes can be significant while it is always negligible for metal fiber brushes. Finally, monolithic brushes are hard and xe2x80x9cbouncexe2x80x9d. At increasing speed, that xe2x80x9cbrush bouncexe2x80x9d must be counteracted by an increasingly strong pressure between brush and substrate at the correspondingly increased friction power loss. Practically speaking, this syndrome limits the sliding speed of monolithic brushes to about 25 m/sec, as already indicated, whereas metal fiber brushes are intrinsically flexible (i.e., have a much larger xe2x80x9cmechanical compliancexe2x80x9d). Therefore, they can and should be mechanically only lightly loaded and can be operated to high speeds at only minor friction heat loss.
Metal foil brushes [3] closely resemble metal fiber brushes except that they are composed not of substantially parallel fibers but of thin parallel foils. Consequently they typically have many fewer, but otherwise the same kind of, contact spots. Thus metal foil brushes are very similar to metal fiber brushes but cannot match their attainable current densities, sliding speeds and low power losses. At any rate, foil brushes are based on the same principle as metal fiber brushes, namely electrical contact to the substrate at a large number of microscopically small, bare metal-metal contact spots, optimally lubricated by a double monomolecular layer of adsorbed water. Hence, also, in terms of number of contact spots per unit working surface area (i.e., xe2x80x9ccontact spot densityxe2x80x9d), and mechanical load per contact spot, exactly the same theory applies to metal foil as to metal fiber brushes [4-6].
As stressed, on account of their different geometry, foil brushes comprise a substantially smaller density of contact spots than well-constructed metal fiber brushes. By way of numerical example, the working surface of a typical metal fiber brush constructed of d=50 xcexcm copper wires of about f=15% packing fraction contains roughly 10,000 contact spots per cm2, namely at the individually flexible fiber ends. In a foil brush with, say, df=25 xcexcm thick parallel foils and f=50% packing fraction, there are about 600 contact spots per cm2, located at the foil edges sliding on the substrate, with an estimated three contact spots per foil edge [3]. Correspondingly, without suitable modifications of the substrate, foil brushes will be very superior to monolithic brushes but fall short of metal fiber brushes.
In the background art, it has long since been recognized that the quality of the substrate surface preparation has a strong impact on brush performance in terms, especially, of electrical resistance and wear rate. The latter is commonly stated in terms of xe2x80x9cdimensionless wear ratexe2x80x9d, xcex94l/L, i.e. brush shortening through wear divided by the sliding path length. Dimensionless wear rates in the low 10xe2x88x9211 range are generally desired, and better of about 10xe2x88x9212. To put this last figure in perspective, consider that even fast running machines will rarely exceed sustained speeds of v=40 m/sec of relative motion between brush and substrate, and that a machine overhaul would probably be necessary after one year, i.e. t=3.15xc3x97107 sec, independent of brush performance. The desired sliding path length is then L=tv=1.26xc3x97109 m. With a dimensionless wear rate of xcex94l/L=10xe2x88x9212 the brush would thus have worn by xcex94l=10xe2x88x9212xc3x971.26xc3x97109 m=1.3 mm between maintenance periods. With a long brush and built-in high mechanical compliance, such brush shortening might well be accommodated without any mechanical forward motion of the current connection between brush and machinery, simply through elastic deformation of the brush body at tolerable decrease of brush force. Simultaneously with this great simplification of brush force application, there would be much less wear debris than for monolithic brushes, especially in view of the typically much higher current densities (i.e., smaller areas of brush working surface), and that only the xe2x80x9cpacking fractionxe2x80x9d of f≅15% of the brush body is occupied by fibers, while the 85% voidage generates no debris. Distinctly less favorable but still highly acceptable would be a 10xe2x88x9211 dimensionless wear rate accompanied by 1.3 cm brush shortening and ten times the wear debris. However, such shortening, and in any event short brushes, would require some mechanical means for advancing the brush as it wears, thereby maintaining the brush force approximately constant, i.e. within a factor of about two or less.
The discussed low dimensionless wear rates are not easily achieved. In fact, wear particles form at contact spots where these momentarily mechanically interlock across the interface (i.e., through a momentary mechanical interlocking of brush and substrate). That this is so has been previously shown by Y. J. Chang and the inventor [7,8] and strong additional support for this fact has been obtained by J. L. Young in an M. S. thesis recently completed under the inventor""s supervision [9]. It follows, then, that wear will be strongly reduced by making the substrate as hard and as smooth as possible. Proposals to do so with the lowest possible loss of electrical conductivity are a large part of the present invention.
Even though the theoretical background outlined above is available in the open literature, with extensive research and theoretical studies on electrical contact spots going back to the outstanding pioneering research by R. Holm [2], no previous directed attempt is known to modify substrates with the particular aim of influencing the number of morphology of contact spots for the purpose of improving electrical brush conduction and/or wear rates, as done herein. In the past it was simply recognized that substrate xe2x80x9crun-outxe2x80x9d (i.e., radial deviations in the course of one revolution), should be kept low and that, before use, substrates should be smoothed with fine emery paper. Further, routinely monolithic brushes contain mild abrasives to xe2x80x9ccleanxe2x80x9d the contact, besides the fact that by itself graphite abrades. However, it seems that in the past only the inventor and co-workers have endeavored to discover the underlying reasons which according to the present invention are based on contact spot behaviors. The only modification of substrate shapes for the improvement of brush performance to ever come to the inventor""s notice, is a spiral groove used in the Westinghouse laboratories that was claimed to counteract aerodynamical lift of monolithic brushes.
The present invention presents methods for the management of the contact spots at brush-substrate interfaces of all three types of electrical brushes. The favorable impact of the management of contact spots according to the present invention may be the greatest on the performance of metal foil brushes which as a result may well become competitive with metal fiber brushes, but is expected to be significant to strong also for metal fiber and monolithic brushes. This management of the contact spots is mainly effected through suitable shaping and otherwise conditioning the substrate surface, and to a lesser extent also through modifications of the brush working surfaces, especially in connection with commutation.